How do mics become out of phase?

[mrmeggs]

If anyone could shed some light on what out of phase is, to reverse phase and the reason why you do it, I would very much appreciate it. I have this idea in my head about what it means but need much more clarification. Cheers.//

 

[Atterion]

In-Phase means the sound waves are the same in both Left and Right Channels. Out of phase means, one signal is opposite to other. Fully out of phase would mean the left signal was 180 degrees opposite to the right. As the waves go out of phase they cancel each other out. This occurs often when using 2 mics because of there placement, and how the sound reaches them (at different times, and different points along the waveform). If you have phase problems when micing, it is usually a quick fix by moving one mic.

 

[boingoman]

I'll give it a shot.

Polarity and phase are different things, usually, in audio. Phase reversal is a misnomer. It should be called polarity reversal.

In this case, out of phase has to do with two audio signals from the same source, one of which is delayed in respect to the other. It can happen with more than two, as well, but two is easy to think about.

A common way for this to happen is two mics capturing the same source from different distances.

When the signals are played back, one of the signals is delayed in relation to the other, as it took longer to reach the farther microphone. The two signals are out of phase when played back together, in other words.

When the two out-of-phase signals are combined, some frequencies are cut and some are boosted, to different degrees. This results in what is called "comb filtering", as the peaks and dips in the combined response of two out of phase signals can resemble the teeth of a comb. It has a characteristic sound, and is used in guitar pedals and signal processors as part of the "phaser" effect. A "swooshy" hollow kind of sound.

Change the distance of one of the mics, different freqs boost and cancel.

Put the mics the same distance away from the source, no cancelling.

Polarity reversal is different, in audio. Plus and minus are reversed. There is no delay involved.
Example:
A signal goes from 0, up to 1, and back to zero.
If the polarity is reversed, the signal goes from 0, to -1, and then back up to zero.
Adding identical reversed polarity signals together results in no sound, in theory. The opposing signals cancel each other out at all points, as for every + value there is a - value at the same time.

This is what happens when you miswire one of a set of stereo speakers. It has the wrong polarity. The low end often disappears, as it is usually in both channels and cancels. Anything else that is in both channels will drop in volume as well. If your system is mono, most everything will disappear.

Another example is micing top and bottom of a snare drum at the same time. The polarity of one must be reversed. Striking the head of the drum results in one mic diaphragm moving one way, the other the opposite way. The two signals are of opposite polarity at the mics. One must be switched to prevent wideband cancellation of the signal on playback.


Now, put the two concepts together. Reversing the polarity of a microphone can help with phase issues by reversing the polarity of frequencies which are cancelling due to being out of phase. Other frequencies cancel, but the problem should be more obvious one way or the other. Polarity reversal (generally) moves a lot of comb filtering out of the range you are trying to fix. (hopefully)

Example:
Your overhead mics are at exactly the right height so that the snare drum sounds like crap when you listen to your snare and overhead tracks together. The snare sound is being affected due to phase issues.
Solutions:
Move the overheads, reverse the polarity of either the overheads or snare, or maybe even do both.

 

[mattamatta]

Pretty good explaination there.

If you want to go deeper or really cement your understanding more, you might look up something from a physics tutorial about waves and oscillation. Phase is a property of anything that repeats cyclically (even light when it's acting as a wave, and some interesting things can happen with it), and there's a lot of study that can be done regarding phase differences in various kinds of waves or cyclical occurrences, including sound, light, and audio signals. In fact, the physics class I'm currently in went all through waves just a few weeks ago. Pretty easy stuff once you get it; I had a great physics class in high school so the concepts make a lot of sense to me.

It really does apply very directly to audio, as the physics of sound waves (or electric signals representing the sound waves) dictates almost everything we deal with as recordists.

 

[Atterion]

When the signals are played back, one of the signals is delayed in relation to the other, as it took longer to reach the farther microphone. The two signals are out of phase when played back together, in other words.

OK the whole explanation above was what I was shooting at. That was a much better explanation boingoman. Except in this quote here. Which vary much is correct, still assumes all signals will be out of phase. It is however very likely for phase alignment even if one signal is delayed (although the chances for perfect alignment is highly unlikely due to room dynamics). Case in point: close mic'd amp, combined with ambient mic. 2 Mics, 2 distances, not necessarily a problem.

One other consideration besides microphone distances from sound source is reflections. If reflected waves (i.e. sounds bouncing of walls, furniture, etc.) reach the microphone at the same time as the source sound, there is potential for phase cancellation.

mrmeggs, if you have doubts about wheter you are experiencing phase problems, utilize a phase-scope (even the most basic audio editors feature one). And double check your mix in mono, which will surely reveal to you whether or not there is a problem.

 

[boingoman]

Which vary much is correct, still assumes all signals will be out of phase. It is however very likely for phase alignment even if one signal is delayed (although the chances for perfect alignment is highly unlikely due to room dynamics). Case in point: close mic'd amp, combined with ambient mic. 2 Mics, 2 distances, not necessarily a problem.

One other consideration besides microphone distances from sound source is reflections. If reflected waves (i.e. sounds bouncing of walls, furniture, etc.) reach the microphone at the same time as the source sound, there is potential for phase cancellation.

Good points.

 

[bennychico11]

Polarity and phase are different things, usually, in audio. Phase reversal is a misnomer. It should be called polarity reversal.

actually, it's more of a word being misspelled than a misnomer. It's technically supposed to be called phase inversing.
People just get lazy and consider inverse and reverse the same thing.

 

[donkeystyle]

actually, it's more of a word being misspelled than a misnomer. It's technically supposed to be called phase inversing.
People just get lazy and consider inverse and reverse the same thing.

actually you may be right about the inverse part, but it's still polarity, not phase. so you may both be partially right.

 

[Atterion]

Actually they're both right, because polarity only refers to the electrical aspect of it. But phases still exist in electricity (Ever heard of 3-Phase wiring?). When we refer to phase, we think of sound waves through the air, but those waves still exist in line form, but are converted from mechanical to electrical form via the microphone, so the new electrical signals are still phase coherent to the original, just in a different form. So inversing the polarity of the line, is essentially the same as reversing the phase prior to it's arrival at the microphone (for the purposes of this discussion anyways. I know there are electrical differences but they are pretty much irrelevant, because In the case of audio, polar orientation is only a factor until a signal is summed to another, and final output (Speakers) are independant of the source (Unless the music is intentionally mixed for 1 speaker's polarity inversed, such as Richard Wright's "Broken China" album, but this is still a seperate process anyways).

 

[mixsit]

actually, it's more of a word being misspelled than a misnomer. It's technically supposed to be called phase inversing.
People just get lazy and consider inverse and reverse the same thing.

Is inverting the word you're looking for?
Wayne

 

[bennychico11]

Is inverting the word you're looking for?
Wayne

yeah yeah, i need to work on my grammar....inverting is the correct tense
of inverse that i was looking for.

 

[boingoman]

Actually they're both right, because polarity only refers to the electrical aspect of it. But phases still exist in electricity (Ever heard of 3-Phase wiring?). When we refer to phase, we think of sound waves through the air, but those waves still exist in line form, but are converted from mechanical to electrical form via the microphone, so the new electrical signals are still phase coherent to the original, just in a different form. So inversing the polarity of the line, is essentially the same as reversing the phase prior to it's arrival at the microphone (for the purposes of this discussion anyways. I know there are electrical differences but they are pretty much irrelevant, because In the case of audio, polar orientation is only a factor until a signal is summed to another, and final output (Speakers) are independant of the source (Unless the music is intentionally mixed for 1 speaker's polarity inversed, such as Richard Wright's "Broken China" album, but this is still a seperate process anyways).

Phase and polarity start diverging when you talk about a complex non-repeating waveforms like music. Phase is only really relevant to simple waves, like a single audio frequency or sine waves.

For a pure sine wave, 180 degrees of phase shift and polarity reversal (or inversion, or whatever) are indistinguishable.

X degrees of phase shift in a complex non-repeating signal means nothing unless it is referenced to a specific frequency.


If you play two identical guitar tracks, one delayed to represent 180 degrees of phase shift at 100 hz, the result will sound very different than delaying one to represent 180 degrees of phase shift at 2500 hz, and neither will sound like polarity reversal.

Phase shift in eqs and speakers means the same thing, but has different implications and results.
EQs and crossovers introduce a certain amount of phase shift to a signal at different frequencies. EQ filters tend to do it broadband, plus extra at the boost/cut freqs, crossovers tend to exhibit phase shift at just certain frequencies.

The difference is this:
Two out of phase tracks played together exhibit cancellation.
A speaker that has severe phase shift at a certain frequency will not exhibit cancellation to the same degree, as it is not played against a signal that has no phase shift in it. It may affect harmonics of the delayed frequencies, though.

It can also sound "smeared" at certain frequencies, as the whole signal is not time-coherent. 45 degrees of phase shift at 12k across your whole mix is gonna sound pretty odd.

You are listening to one signal with certain freqs delayed, rather than two signals that are delayed with respect to each other.

 

[bennychico11]

the way i see it, both words represent exactly what we want it to do.
In the mixer, the polarity reversal button switches polarity of the signal and in result switching the rise and fall of the wave form. When we speak of phase inverting, we mean we want the signal to be 180 degree from where it is now....which is meaning switching the rise and fall of the wave form.

polarity reversal happens in the electronics of it all
phase inverting happens in between the speakers and our ears.
they both have the same end result.....switching the wave form 180 degrees around.

 

[boingoman]

The difference is slight, but important.
Phase refers to a time lead/lag.
Polarity refers to time coincident signals that have the sign of their amplitude values reversed.

In a simple repeating wave, they can mean the same thing. In a complex non-repeating, non-symmetrical waveform, they cannot. Phase shift which causes cancellation at one frequency won't cause cancellation at all frequencies.

 

[bennychico11]

The difference is slight, but important.
Phase refers to a time lead/lag.
Polarity refers to time coincident signals that have the sign of their amplitude values reversed.

In a simple repeating wave, they can mean the same thing. In a complex non-repeating, non-symmetrical waveform, they cannot. Phase shift which causes cancellation at one frequency won't cause cancellation at all frequencies.

agreed
but we're not talking about phase in general. we're talking about phase inverting on an entire signal on all frequencies (which is what happens when you press that little Theta button on the mixer). and inverting the phase on that signal (which in the electrical world is switching polars) will give us the total opposite of the original. And if the two were played at the exact same amplitude at the exact same time, then they would cancel eachother out (theoretically).

i guess people just like to say "Switch the phase" when they want to change what they're hearing in the acoustical world. it's a lot like when people say "turn up the volume." Yes you're turning up the volume, but you're also turning the variable resistor. Electrically that's what we're doing...but who cares what's going on in the electronics at the moment. We want to hear what's going on in our world that our ears are listening to.

...but maybe we should just start saying that.
PUMP UP THE VARIABLE RESISTOR!!!

 

[boingoman]

agreed
but we're not talking about phase in general. we're talking about phase inverting on an entire signal on all frequencies (which is what happens when you press that little Theta button on the mixer). and inverting the phase on that signal (which in the electrical world is switching polars) will give us the total opposite of the original. And if the two were played at the exact same amplitude at the exact same time, then they would cancel eachother out (theoretically).:

LOL. True enough. Polarity change results in phase inversion at all frequecies.
I think I just am trying to point out the difference between adjusting phase using delay or different mic distances vs. polarity reversal.

...but maybe we should just start saying that.
PUMP UP THE VARIABLE RESISTOR!!!

Or my favorite, "Turn it down less!!"

 

[Atterion]

we're talking about phase inverting on an entire signal on all frequencies (which is what happens when you press that little Theta button on the mixer).

This was the point I was trying to make earlier. Of course phase cancellation is frequency dependent. The vary fundamental of a waveform is its frequency, so obviously a signal can only only be fully cancelled by an opposite phase waveform of the same frequency, and can be partially affected to varying degrees by any waveform of any frequency. What I was saying is reversing the polarity of the line, inverts all waveforms present on that line, complex, basic or otherwise (Of course this assumes that every component of the circuit is operating at an unrealistic 100% efficiency, and offering no impedence. Of course they don't so line-phase anomilies are created. This is what I hinted at earlier).

 

[Atterion]

I think I just am trying to point out the difference between adjusting phase using delay or different mic distances vs. polarity reversal

That is fully understood. And is the preferred method. To get it right in the first place. But polarity, is still important, and shouldn't be dismissed. It's a part of the puzzle, wheter we choose to acknowledge it or not. Have we lost you yet mrmeggs???

 

[boingoman]

Have we lost you yet mrmeggs???

I was wondering that myself.

 

[Atterion]